Natural menopause timing and menopause symptoms among lesbian, bisexual, and heterosexual women in the Nurses’ Health Study II

Abstract

Objective:

This study aimed to examine sexual orientation differences in natural menopause timing and symptoms between lesbian and bisexual women compared with heterosexual women.

Methods:

We used longitudinal questionnaire data (1989–2015) from 92,314 women (858 lesbian, 375 bisexual) in the Nurses’ Health Study II cohort. Women were 24–44 yr old at baseline and biennially reported their menopause status, including reasons for cessation of menstrual periods. In 2009 and 2013, women reported on their experience of hot flashes and night sweats. Covariates included age, age at menarche, body mass index, smoking, and parity. Sexual orientation was reported in 1995 and 2009.

Results:

Age-adjusted Cox models and logistic regression models suggest that there were no statistically significant differences in menopause timing between lesbian or bisexual women when compared with heterosexual women. When examining menopause symptoms, lesbian women consistently had a statistically significantly (P < 0.05) higher odds of experiencing hot flashes or night sweats (odds ratio range: 1.17 to 1.72) and moderate/severe symptoms (OR range: 1.26 to 1.77) than heterosexual women, even after adjusting for covariates such as smoking and obesity. There were no statistically significant differences in menopause symptoms between bisexual and heterosexual women.

Conclusions:

Our findings suggest no meaningful difference in menopause timing between sexual minority and heterosexual women. Additionally, menopause symptoms were more likely among lesbian women and warrants additional study.

Younger age of menopause can increase risk for cardiovascular disease, osteoporosis, cognitive decline, and premature mortality.^1–8 Further, symptoms experienced during menopause, such as hot flashes and night sweats, can negatively impact quality of life.^9,10 Numerous lifestyle and behavioral factors, such as cigarette smoking, parity, stress, and diet, as well as biochemical factors, such as hormone profiles, may affect menopause timing and symptoms.

Lesbian and bisexual women are more likely than heterosexual women to experience more obesity, smoking, excessive alcohol use, depression, and abuse or victimization.^11–22 These factors have been linked to adverse timing of menopause as well as menopause symptoms at midlife.^23–29 Minority stress theory is the primary guiding framework for explaining poorer health among sexual minority women. Minority stress posits that the social discrimination, negative bias, and stigma imposed on sexual minorities negatively affect their mental and physical health.^30,31 Stress may have a direct effect on menopause timing or symptoms as well as an indirect effect. For example, it can be hypothesized that due to the added social stress of being a sexual minority in society, sexual minorities may exhibit more depression or anxiety as well as riskier health behaviors (eg, cigarette smoking), which puts them at greater risk for subsequent disease.

Collectively, this existing evidence suggests that lesbian and bisexual women exhibit more risk factors than heterosexual women for younger age at menopause and greater symptom experience than heterosexual women. However, little research has examined sexual orientation differences in menopause. We identified only four studies that examined sexual orientation disparities in menopause, all of which were cross-sectional and assessed timing only. Three studies,^32–34 all based in California, examined menopause, but findings are limited due to extensive missing data regarding menopause,³² examination of menopause status (ie, premenopause, perimenopause, or postmenopause) rather than menopause timing,³³ and small sample sizes (n lesbian = 30),³⁴ which could explain null findings. Most notably, none of these studies included bisexual women, which is a substantial limitation because even less is known about bisexual women’s menopause health. The fourth study, which included lesbian and bisexual women, cross-sectionally examined a wave of data from our cohort, the Nurses’ Health Study II (NHSII), when participants were 32–51 yr of age.¹⁷ However, at the time of the study, only 5% of study participants had experienced menopause. We were unable to identify any epidemiologic studies examining sexual orientation differences in menopause symptoms.

To address this gap in research, we used all available waves of NHSII (nearly 30 yr of data) to 1) prospectively assess sexual orientation differences in menopause timing and 2) evaluate sexual orientation differences in experiencing menopause symptoms, including hot flashes and night sweats, as well as the severity of these symptoms. We hypothesized that compared with heterosexual women, lesbian and bisexual women would experience menopause at younger ages and would report experiencing more menopause symptoms, as well as greater symptom severity, even after accounting for important behavioral factors.

METHODS

Study population

The NHSII is an on-going longitudinal cohort of 116,430 female registered nurses who were 24–44 yr old in 1989 and responded to a mailed questionnaire. Information on medical conditions and behavioral health is assessed on biennial questionnaires. Follow-up rates over nearly 30 yr have been excellent (85%−90%). The study protocol was approved by the Institutional Review Board at Brigham and Women’s Hospital in Boston, Massachusetts.

Sexual orientation

Sexual orientation was assessed in 1995 and 2009, with participants choosing from the following response options: “heterosexual,” “lesbian, gay, or homosexual,” “bisexual,” “none of these,” and “prefer not to answer.” Consistent with previous work using these data,^35–37 sexual orientation reported in 2009 was used, except where information was missing, in which case sexual orientation reported in 1995 was used. Participants who responded “none of these” or “prefer not to answer” in both 1995 and 2009 were excluded from analyses.

Age at menopause

Beginning on the 1989 baseline questionnaire and on all follow-up questionnaires, participants were asked if their periods had ceased permanently with the following response options: “no: premenopausal,” “yes: no menstrual periods,” “yes: had menopause but now have periods induced by hormones,” and “not sure (eg, started hormones before the cessation of periods.” Those who indicated that their periods had ceased were asked the following questions: 1) At what age did your periods cease? (open response); and 2) For what reason did your periods cease? (response options: “surgery,” “radiation or chemotherapy,” and “natural”). Consistent with previous work using these data, age at natural menopause was defined as age after 12 consecutive months of amenorrhea that was not due to radiation, chemotherapy, or surgery.³⁸

Menopause symptoms

Using a measure developed for NHS, in 2009 and 2013, participants reported whether they experienced hot flashes or night sweats at the beginning of menopause and in the past 4 wk. Participants who reported experiencing hot flashes or night sweats either at the beginning of menopause or in the past 4 wk were additionally asked about the severity of symptoms (response options: “mild,” “moderate,” or “severe”) during that time. We combined the severity categories of “moderate” and “severe” into a single “moderate/severe” category and compared this group of participants to those who reported no menopause symptoms. In 2009, among those who reported menopause symptoms at the beginning of menopause, 59.9% also reported symptoms in the past 4 wk. This number dropped to 50.9% in 2013. Due to potential recall bias, particularly among participants years beyond menopause, responses for menopause symptoms at the beginning of menopause in 2013 were excluded from analysis.

Covariates

Covariates included age, age at menarche (assessed at baseline), as well as smoking history in pack-years, parity, hormone therapy (HT), and body mass index, which were assessed multiple times during follow-up. We assessed sociodemographic factors, such as region of residence, rural living status, income, and race/ethnicity, as covariates; because these factors did not substantially change findings, we did not include them in these results.

Statistical analysis

Women were excluded from analyses if they were missing sexual orientation information in both 1995 and 2009, reported a hysterectomy or oophorectomy, history of cancer (not including non-melanoma skin cancer), or were postmenopausal at baseline. After baseline exclusions, the final eligible sample size was 92,314. In all analyses, heterosexual women were the referent group. All analyses were conducted in SAS version 9.4 (SAS Institute Inc).

To assess sexual orientation differences in age at menopause, we used Cox proportional hazards models. Follow-up for all participants began from baseline until menopause, hysterectomy, oophorectomy (bilateral or unilateral), cancer diagnosis (not including non-melanoma skin cancer), loss to follow-up, or death, whichever occurred first. To examine whether lesbian and bisexual women were more likely to experience hot flashes or night sweats than heterosexual women, we fit logistic regression models. We also fit logistic regression models to assess the severity of symptoms (moderate/severe vs no symptoms). Both unadjusted and adjusted models were fit, with adjusted models including age, age at menarche, smoking, body mass index, parity, and HT use (for symptom analyses only) as covariates.

Sensitivity analyses

We performed several additional sensitivity analyses to better contextualize and understand primary study findings. First, we fit the same models for age at menopause and menopause symptoms but used a more comprehensive multidimensional sexual orientation variable (Supplemental Tables 1, http://links.lww.com/MENO/B310 and Supplemental Table 2, http://links.lww.com/MENO/B311, respectively). A growing body of research highlights the importance of collecting and utilizing multidimensional measures of sexual orientation.³⁹ Indeed, using a multidimensional measure allows for a more comprehensive understanding of sexual minority health by taking into account not just sexual identity, but also attraction and sexual contact. In 2017, additional dimensions of sexual orientation were assessed by capturing participants’ identities, attractions, and sexual contacts. Current identity was measured with the following response options: “completely heterosexual,” “mostly heterosexual,” “bisexual,” “mostly homosexual,” “completely homosexual,” and “not sure.” Participants also reported whether they have ever: identified as mostly heterosexual, bisexual, or lesbian or gay; been sexually attracted to males; and been sexually attracted to females. Participants reported having sexual contact with: “male(s),” “female(s),” “male(s), and female(s),” and “I have not had sexual contact with anyone.” Using all available sexual orientation data from 1995, 2009, and 2017, we created five sexual orientation categories: completely heterosexual (identified as completely heterosexual and never had past same-sex attractions/partners nor identified previously as a sexual minority); heterosexual with past same-sex attractions/partners/sexual minority identity; mostly heterosexual; bisexual; and lesbian. Second, we examined the association between sexual identity and menopause symptoms (Supplemental Table 3, http://links.lww.com/MENO/B312) but restricted analysis to women between 45–55 yr of age in 2009 and 2013 to ensure that findings were not substantially biased by older women who may be well past menopause. Third, we examined the association between sexual identity and menopause symptoms but excluded those taking HT (Supplemental Table 4, http://links.lww.com/MENO/B313).

RESULTS

Throughout follow-up, 73,456 (80%) women experienced natural menopause. At baseline, lesbian (n = 858) and bisexual (n = 375) women were more likely to smoke cigarettes (0 pack-years: lesbian = 49%; bisexual = 49%; heterosexual = 66%), have higher body mass index (lesbian = 25.2 kg/m²; bisexual = 25.3 kg/m²; heterosexual = 23.9 kg/m²), and be nulliparous (lesbian = 74%; bisexual = 49%; heterosexual = 28%) than heterosexual women (n = 91,081).

Sexual orientation differences in timing of menopause are presented in Table 1. Unadjusted results suggested that there were no differences in timing of menopause between lesbian [HR (95% CI): 1.05 (0.97–1.13)] and bisexual women [0.94 (0.84–1.06)] compared with heterosexual women. After adjusting for age at menarche, smoking, body mass index, and parity, there remained no difference between lesbian and heterosexual women [0.98 (0.91–1.06)] or between bisexual and heterosexual women [0.91 (0.81–1.02)].

TABLE 1.

Cox proportional hazards regression models estimating sexual orientation differences in menopause timing in the Nurses’ Health Study II, 1989–2015

	Heterosexual	Lesbian	Bisexual
Cases/person-years	72,458/1,308,301	707/11,687	292/5,221
	Hazards Ratio (95% Confidence Interval)
Model 1a	1.00 (ref)	1.05 (0.97 to 1.13)	0.94 (0.84 to 1.06)
Model 2b	1.00 (ref)	0.98 (0.91 to 1.06)	0.91 (0.81 to 1.02)

^aAdjusted for age only.

^bAdjusted Model 1 covariates and age at menarche (continuous), smoking (in pack-years; 0, 1–5, 6–10, 11–15, 16–20, >20), body mass index (in kg/m²; continuous), and parity (nulliparous, 1–2, or >2).

Sexual orientation differences in menopause symptoms are presented in Table 2. Compared with heterosexual women, lesbian women generally had statistically significantly higher odds of experiencing hot flashes or night sweats (OR range: 1.17 to 1.72, P < 0.05); however, there were no differences in experiencing menopause symptoms between bisexual and heterosexual women. Among those who reported experiencing menopause symptoms, there was generally no statistically significant difference in symptom severity by sexual orientation.

TABLE 2.

Unadjusted and adjusted logistic regression menopause symptom differences between lesbian and bisexual compared with heterosexual women, Nurses’ Health Study II, 2009 and 2013

		OR (95% CI)
	2009			2013
	Heterosexual	Lesbian	Bisexual	Heterosexual	Lesbian	Bisexual
Beginning of menopause
Hot flashes or night sweats (yes vs no)
	n = 55,793	n = 599	n = 241
Model 1a	1.00 (ref)	1.65 (1.32 to 2.08)***	0.97 (0.71 to 1.31)
Model 2b	1.00 (ref)	1.70 (1.35 to 2.14)***	0.99 (0.73 to 1.34)
Model 3c	1.00 (ref)	1.72 (1.37 to 2.17)***	0.99 (0.73 to 1.34)
Severity (moderate/severe vs no)
	n = 40,824	n = 439	n = 188
Model 1a	1.00 (ref)	1.73 (1.37 to 2.19)***	1.05 (0.76 to 1.43)
Model 2b	1.00 (ref)	1.72 (1.36 to 2.19)***	1.04 (0.76 to 1.43)
Model 3c	1.00 (ref)	1.77 (1.40 to 2.26)***	1.05 (0.76 to 1.44)
Past 4 wk
Hot flashes or night sweats (yes vs no)
	n = 70,743	n = 711	n = 294	n = 75,465	n = 739	n = 303
Model 1a	1.00 (ref)	1.18 (1.02 to 1.37)*	0.96 (0.77 to 1.21)	1.00 (ref)	1.30 (1.12 to 1.50)***	1.04 (0.83 to 1.31)
Model 2b	1.00 (ref)	1.18 (1.02 to 1.37)*	0.97 (0.77 to 1.22)	1.00 (ref)	1.32 (1.14 to 1.53)***	1.06 (0 to 1.33)
Model 3c	1.00 (ref)	1.19 (1.02 to 1.38)*	0.96 (0.76 to 1.21)	1.00 (ref)	1.32 (1.14 to 1.54)***	1.06 (0.84 to 1.33)
Severity (moderate/severe vs no)
	n = 51,018	n = 524	n = 213	n = 55,123	n = 529	n = 225
Model 1a	1.00 (ref)	1.37 (1.15 to 1.63)***	0.96 (0.71 to 1.28)	1.00 (ref)	1.42 (1.17 to 1.71)**	1.11 (0.82 to 1.51)
Model 2b	1.00 (ref)	1.30 (1.08 to 1.55)**	0.92 (0.69 to 1.24)	1.00 (ref)	1.34 (1.11 to 1.63)**	1.08 (0.80 to 1.46)
Model 3c	1.00 (ref)	1.30 (1.09 to 1.56)**	0.92 (0.68 to 1.23)	1.00 (ref)	1.35 (1.11 to 1.63)**	1.08 (0.80 to 1.46)

^*P < 0.05,

^**P < 0.01,

^***P < 0.001.

^aAdjusted for age.

^bAdjusted for Model 1 covariates and age at menarche (continuous), smoking (in pack-years; 0, 1–5, 6–10, 11–15, 16–20, >20), body mass index (in kg/m²; continuous), and parity (nulliparous, 1–2, or >2).

^cAdjusted for Model 2 covariates and hormone therapy use (current user, past user ever, past user this survey cycle, never user, missing).

Sensitivity analyses

Findings from sensitivity analyses using a multidimensional measure of sexual orientation were similar to primary analyses using sexual identity; however, importantly, for experience of menopause symptoms, some of the sexual minority women defined by the more expansive sexual identity options and the addition of sexual attraction and contact were slightly more likely to report experiencing hot flashes or night sweats and greater severity (Supplemental Table 1, http://links.lww.com/MENO/B310 and Supplemental Table 2, http://links.lww.com/MENO/B311). However, because collection of multidimensional sexual orientation occurred after participants experienced menopause, we opted for prior measures of sexual identity only, which occurred concurrently with menopause assessment. Findings from analyses of sexual identity and menopause symptoms using restricted samples (Supplemental Table 3, http://links.lww.com/MENO/B312: limited to 45–55 and Supplemental Table 4, http://links.lww.com/MENO/B313: excluding HT users) were similar to findings from the full sample. However, smaller sample sizes resulted in wider confidence intervals and less precise estimates.

DISCUSSION

This work is an important first step in understanding sexual orientation disparities in menopause health, a highly understudied area of women’s health, and lesbian and bisexual women’s health in particular. This work can help us more effectively prioritize and develop public health, health care, and research efforts to address these disparities. In line with this, our findings suggest that there may not be substantive differences in age at menopause between lesbian, bisexual, and heterosexual women. However, important differences may exist in terms of the experience of hot flashes and night sweats between lesbian and heterosexual women.

This initial investigation into timing of menopause and experience of menopause symptoms using nearly 30 yr of prospective cohort data highlights several important areas for consideration. First, lesbian women may experience menopause differently than heterosexual women, notably with regard to hot flashes and night sweats. Although age at menopause did not differ substantially between lesbian and heterosexual women, lesbian women were much more likely to report experiencing hot flashes or night sweats and to report these experiences as being moderate or severe compared with heterosexual women. This difference in reports of menopause symptoms between lesbian and heterosexual women, while modest, was consistent even after adjusting for important covariates associated with earlier menopause timing, including smoking history as well as in both time periods when menopause symptoms were assessed. Additional exploration into what may contribute to these differences may help inform future interventions to help lesbian women more effectively manage their menopause symptoms.

Several limitations need to be considered. First, the NHSII study population is comprised of professional nurses who are primarily non-Hispanic White. Therefore, findings may not generalize to women of color or women in low or high socioeconomic positions. Second, we relied on self-report of menopause status to determine timing of menopause, which could lead to some misclassification of menopause timing. The likelihood of misclassification is low given that a previous study in a similar study population found that self-report of menopause was found to be valid and reliable.⁴⁰ Although we were able to examine timing of menopause prospectively, data for experience of menopause symptoms were not as consistently collected; therefore, examination into trajectories of menopause symptom experience during midlife and how they might differ by sexual orientation was not possible. Similarly, menopause symptoms assessment relied on participant recall of experiences and could be subject to recall bias, although past 4-wk assessment should be less impacted by recall. Additionally, we only examined hot flashes and night sweats; other menopause symptoms (eg, joint pain, sleep disturbance, affective symptoms) that may vary by sexual orientation and are associated with lower quality of life were not assessed. Future work examining a variety of menopause symptoms among sexual minority women should use more frequent assessments over multiple time points. Finally, future work should utilize multidimensional measures of sexual orientation. Findings from sensitivity analyses suggest that sexual minority women, defined more broadly than identifying as lesbian or bisexual, may experience menopause symptoms differently than completely heterosexual women. Additional concurrent examination of sexual orientation and menopause symptoms may help better elucidate important subgroup differences.

CONCLUSIONS

Findings provide critical initial evidence on important sexual orientation differences in menopause symptom experience, but not timing, especially for lesbian women. Future work using more diverse longitudinal samples of women as well as more frequent and comprehensive measures of menopause symptoms are needed to corroborate our findings. Additionally, further examination into how and why menopause symptom experience may be greater among lesbian women is needed.